Rule Based Systems for Classification in Machine Learning Context

The aim of this project is to develop a theoretical framework for design of rule based systems for classification tasks in machine learning context. The objectives are:
1. To develop novel methods/techniques relating to rule based classification for advancing rule learning and interpretation.
2. To develop novel frameworks for advancing ensemble learning.
3. To validate the methods/techniques and frameworks described in Objectives 1 and 2.

Theoretical Significance:
-Development of a unified framework for design of rule based systems
-Development of more advanced ensemble learning frameworks
-Development of novel approaches for generation, simplification and representation of rules
-Novel applications of graph theory and BigO notation

Practical Importance:
-Knowledge discovery and predictive modelling
-Parallel, distributed and mobile data modelling
-Domain independent applications

Methodological Impact:
-Complementing existing rule learning methods
-Collaborating with existing rule learning methods
-Advancing interpretability of rule based models

Philosophical Aspects:
-Novel understanding of data mining, machine learning and their difference
-Novel understanding of ensemble learning in the context of learning theory
-Philosophical inspiration from information theory, system theory and control theory